AI-Ready Data Blueprints From Raw Data to AI-Driven Innovation (Navnit Shukla, Kien Pham, Srikanth Sopirala etc.)（Z-Library）

Page 1

Navnit Shukla, Kien Pham, Srikanth Sopirala & Harsha Tadiparthi Foreword by Ehsan Hoque AI-Ready Data Blueprints From Raw Data to AI-Driven Innovation

Page 2

9 7 9 8 3 4 1 6 3 1 7 9 3 5 7 9 9 9 US $79.99 CAN $99.99 DATA ISBN: 979-8-341-63179-3 Companies innovating with generative AI understand that having the right data foundation is critical for success and profitability. To best position themselves for long-term success, organizations must prioritize investments in data and AI governance. AI-Ready Data Blueprints is your map to connecting data strategy, GenAI, and ethical practices to build and scale truly effective solutions. Taking a comprehensive, cloud-agnostic approach focused on real-world business challenges, seasoned data and AI experts Navnit Shukla, Kien Pham, Srikanth Sopirala, and Harsha Tadiparthi share actionable insights to guide you in designing and implementing effective data-centric GenAI systems. Whether you’re new to GenAI or are already focusing on optimizing it for accuracy, speed, or both, the principles shared in this book will empower you to excel in all your AI endeavors. • Identify the key elements of a solid data foundation for generative AI • Apply data governance and orchestration techniques to ensure high data quality, access control, and proper data lineage for reliable AI systems • Optimize GenAI applications through prompt engineering, fine-tuning, and retrieval-augmented generation • Implement security, compliance, and governance measures, including responsible AI practices, transparency, and more Navnit Shukla is a Snowflake principal solutions architect who helps clients derive insights from data using AI. He is also the author of Data Wrangling on AWS. Kien Pham is an AWS principal solutions architect supporting digital native business. Kien has over 10 years of software engineering experience. Srikanth Sopirala is a principal AI specialist at AWS, helping global enterprises turn high-stakes AI and data challenges into secure, scalable, production-ready solutions. Harsha Tadiparthi is a principal AI specialist at AWS, where he helps Fortune 500 companies solve complex challenges in data and AI. AI-Ready Data Blueprints “This book turns the hardest problem in enterprise GenAI—making your data actually AI-ready—into a clear, actionable engineering plan. If you’re building, not just talking, this is your playbook.” Magesh Varadharajan, engineering leader, Gen Digital Inc. “A rare blend of strategy and hands-on execution. A practical guide for translating AI hype into enterprise reality.” Iskander Sanchez-Rola, AI and innovation leader

Page 3

Navnit Shukla, Kien Pham, Srikanth Sopirala, and Harsha Tadiparthi Foreword by Ehsan Hoque AI-Ready Data Blueprints From Raw Data to AI-Driven Innovation

Page 4

979-8-341-63179-3 [LSI] AI-Ready Data Blueprints by Navnit Shukla, Kien Pham, Srikanth Sopirala, and Harsha Tadiparthi Copyright © 2026 Navnit Kumar Shukla, AZ25 Lab, Harsha Tadiparthi, and Srikanth Sopirala. All rights reserved. Published by O’Reilly Media, Inc., 141 Stony Circle, Suite 195, Santa Rosa, CA 95401. O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are also available for most titles (https://oreilly.com). For more information, contact our corporate/institu‐ tional sales department: 800-998-9938 or corporate@oreilly.com. Acquisitions Editor: Aaron Black Development Editor: Sara Hunter Production Editor: Elizabeth Faerm Copyeditor: Rachel Wheeler Proofreader: Kim Wimpsett Indexer: Judith McConville Cover Designer: Susan Brown Cover Illustrator: Monica Kamsvaag Interior Designer: David Futato Interior Illustrator: Kate Dullea May 2026: First Edition Revision History for the First Edition 2026-05-06: First Release See https://oreilly.com/catalog/errata.csp?isbn=9798341631793 for release details. The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. AI-Ready Data Blueprints, the cover image, and related trade dress are trademarks of O’Reilly Media, Inc. The views expressed in this work are those of the authors and do not represent the publisher’s views. While the publisher and the authors have used good faith efforts to ensure that the information and instructions contained in this work are accurate, the publisher and the authors disclaim all responsibility for errors or omissions, including without limitation responsibility for damages resulting from the use of or reliance on this work. Use of the information and instructions contained in this work is at your own risk. If any code samples or other technology this work contains or describes is subject to open source licenses or the intellectual property rights of others, it is your responsibility to ensure that your use thereof complies with such licenses and/or rights.

Page 5

Table of Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1. Introduction to AI-Ready Data Foundation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction and Market Context 2 What Makes Generative AI Different 4 The Transformer Architecture: The Technical Foundation of GenAI 5 Enterprise Data as the Key Differentiator 6 Contextual Intelligence: A Simple Example 7 Representing Meaning in Vector Space 8 Enterprise Example: Cross-Document Understanding in Customer Support 10 The Evolution of GenAI Applications 12 From Assistants to Agents: The Four Stages of GenAI Evolution 12 The Increasing Complexity of Data Requirements 15 What Leading Organizations Are Building Today 16 The Production Reality Check 16 Emerging GenAI Architectural Patterns 17 From Patterns to Practice 22 Development Cycles: ML Versus GenAI 23 Traditional Machine Learning Development Cycle 23 GenAI Development Cycle 26 Data Infrastructure Implications 29 Architecture Evolution in Practice: From Traditional ETL to GenAI-Ready Pipelines 30 Key Differences Between Traditional and GenAI Data Foundations 31 Real-World Example: Evolving from Kimball to Medallion to GenAI-Ready Architecture 32 iii

Page 6

Preparing for the Agent-Driven Future 37 “Jeeves Does the Shopping”: The Agent-Mediated Consumer Experience 38 Search Engine Optimization: From Keywords to Agent Optimization 38 Agents Have No Allegiance: Preparing for Radical Transparency 39 Business Autopilot: Autonomous Operations and Decision Making 40 Data Readiness Checklist for the Agent-Driven Future 41 Blueprint: Implementation Guidance 42 Where to Start 42 The Decision Tree: Which Pattern First? 43 Your Action Plan 44 Summary 45 2. Data Framework for GenAI and Agentic AI Applications. . . . . . . . . . . . . . . . . . . . . . . . . . 47 Introduction: Building the Foundation for AI-Ready Data 48 The Evolution of Data Frameworks 49 Early Days (2015–2019) 49 Transition Period (2020–2022) 50 GenAI Era (2022–Present) 50 Agentic AI Era (2024–Present) 51 The Need for a New Approach: Core Requirements for AI-Ready Data 52 Capturing Business Logic and Context 53 Ensuring Data Quality and Consistency 54 Managing Complexity and Diversity 54 Maintaining Security, Compliance, and Privacy 55 Enabling Information Sharing and Collaboration 55 Supporting Scale and Performance 55 Managing Data as a Strategic Product 56 Empowering Users with Documentation and Guidance 56 A Core Framework for AI-Ready Data 56 Capturing Business Logic and Context 57 Ensuring Data Quality and Consistency 61 Managing Complexity and Diversity 65 Maintaining Security, Compliance, and Privacy 69 Enabling Information Sharing and Collaboration 73 Supporting Scale and Performance 76 Managing Data as a Strategic Product 79 Empowering Users with Documentation and Guidance 83 AI-Ready Data Blueprints for the Data Framework: Practical Implementation Guide 86 Blueprint 1: Business Context Intelligence Engine 86 Blueprint 2: Adaptive Data Quality Orchestration 88 Blueprint 3: Orchestrating Data Diversity and Complexity 89 iv | Table of Contents

Page 7

Blueprint 4: Security-First AI Data Platform 91 Summary 93 3. Data Wrangling and Data Preparation for GenAI and Agentic AI Applications. . . . . . . . 95 The Enterprise Challenge 95 Purpose and Audience 96 The Market Imperative 96 The Strategic Advantage 97 Transformation Journey Preview 98 Understanding the Paradigm Shift 98 The Established Machine Learning Data Pipeline 99 The Tabular Data Paradigm and Its Constraints 99 The Conceptual Revolution in Data Processing 100 New Processing Paradigms and Technical Requirements 101 The Business Impact of the Paradigm Shift 102 Self-Assessment: Where Is Your Organization Today? 103 Building Blocks of GenAI-Ready Data 104 Semantic Understanding Fundamentals 104 Knowledge Graphs and Relationship Modeling 106 Real-Time Data Processing Requirements 107 GenAI Data Preparation Maturity Model 108 Case Study: Retail Organization Transformation 110 The Semantic Layer Architecture 111 Architectural Overview and Core Principles 111 Data Foundation Layer: Building on Lakehouse Architecture 113 Metadata and Ontology Management: Creating Semantic Understanding 114 Transformation and Enrichment Pipeline: Adding Intelligence to Data 115 Vectorization and Indexing Infrastructure: Enabling Semantic Search 116 APIs and Reasoning Layer: Enabling AI Consumption 117 Component Interactions and Data Flows 118 Implementation Considerations and Common Patterns 119 AWS Reference Architecture and Implementation 121 AWS Ecosystem Overview for GenAI Data Preparation 121 Key AWS Services and Capabilities 123 Integration Patterns and Service Selection 125 Governance, Quality, and Observability 126 Expanded Quality Dimensions for GenAI Data 126 Governance Framework Essentials 129 A Governance and Quality Monitoring Framework 130 Monitoring and Observability Approaches 133 Regulatory Compliance and Assurance 133 Case Study: Financial Services Governance Implementation 134 Table of Contents | v

Page 8

Advanced Topics: Agentic AI Data Requirements 135 Agentic AI Fundamentals and Data Implications 136 Enterprise Implementation Examples 137 The Future: Autonomous and AI-Assisted Data Wrangling 138 AI-Powered Data Discovery and Intelligent Classification 138 Conversational Data Preparation and Natural Language Interfaces 139 Autonomous Quality Assurance and Self-Healing Systems 139 Future Directions and Research Frontiers 139 Action Plan 140 Key Principles 140 Immediate Next Steps by Maturity Level 141 Long-Term Strategic Considerations 142 Summary 142 4. Data Governance, Security, Compliance, and Orchestration for GenAI. . . . . . . . . . . . . 143 Data Governance and Data Security for AI Applications 144 The AI Data Governance Operating Model 145 Differences in Governance of Structured and Unstructured Data 148 Data Stewardship and Metadata Management 148 Making Data Accessible for Humans and Agentic AI 148 Testing Our Tools 151 Data Quality 153 Responsible AI 155 Fairness 156 Transparency 158 Accountability 162 Privacy and Security 163 Reliability 164 Human Oversight 165 Data Privacy and Security for GenAI Applications 166 Sensitive Information Protection 167 Topic Restriction and Word Filtering 169 Filtering Tools 171 End-to-End Data Protection 174 LLMOps: AI Workflow Orchestration 175 Orchestration Patterns for AI Applications 175 Agentic Patterns 180 Summary 183 5. Knowledge Bases and Vector Databases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 The GenAI Data Challenge 185 Knowledge Bases: Data Organization and Storage 186 vi | Table of Contents

Page 9

Vector Databases: Data Representation and Similarity Search 186 Retrieval-Augmented Generation: Data Retrieval and Context Assembly 187 Why These Technologies Are Essential 187 Knowledge Base Fundamentals: From Data to Knowledge 188 Data Architecture for Knowledge Bases 188 Data Preparation Pipeline for Knowledge Bases 191 Data Types and Management Strategies 193 Data Quality: The Make-or-Break Factor 197 Vector Database Fundamentals: Data Representation 198 Embeddings and Data Quality 199 Vector Database Infrastructure 200 Open Source Solutions 201 Data Indexing Algorithm Comparison 202 Data Compression Techniques 203 Data Governance Feature Comparison 203 Selecting Based on Data Requirements 204 RAG: The Data Retrieval Layer 206 Why RAG? The Data Access Problem 206 The RAG Architecture: Data Flow 206 Data Chunking: Critical for RAG Performance 209 Data Efficiency and Cost Optimization 216 End-to-End Data Flow: KB → Vector DB → RAG 221 Summary 225 6. AI Application Optimization for Production Readiness. . . . . . . . . . . . . . . . . . . . . . . . . . 227 The Journey from Prototype to Production 227 Preparing Your AI Application for Production Success 229 The Twin Pillars of AI Optimization 229 Engineering Efficient Data Workflows for Production AI Systems 230 Data Quality Assessment and Preprocessing Fundamentals 230 Understanding Context Windows and Their Implications 233 Handling Large Input Contexts 234 Optimizing Inference Quality Using Automated Reasoning 235 Metadata Readiness for Agentic AI Systems 242 The Challenges of Scale and Complexity 242 Pattern: Leveraging Raw Files Directly 243 Building Data-Aware Agents Through Intelligent Semantic Metadata 244 Example Implementation: Leveraging Raw Files Directly 247 From Raw Content to Semantic Intelligence: Constructing the Metadata Layer 252 Key Capabilities for Production Readiness for Agentic AI Platforms 256 Leading AI Agent Platform Choices 257 Table of Contents | vii

Page 10

Core Runtime and Deployment Strategy 258 Framework, Tool Integration, and Protocol Support 259 Memory and Knowledge Management Capabilities 259 Security and Compliance Features 260 Observability and Quality Assurance Support 260 Summary 261 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 viii | Table of Contents

Page 11

Foreword In my lab at the University of Rochester, we spent over a decade building AI systems that listen to a patient’s voice and watch their facial movements to detect early signs of Parkinson’s disease and autism, often before a clinician ever sees them. The models we built were sophisticated. The algorithms were sound. But the hardest problem was never the model. It was the data. We learned this lesson the way most researchers do: painfully. Our early systems would perform beautifully on curated datasets and then fall apart in the real world— not because the neural networks were wrong, but because the data feeding them was incomplete, inconsistent, or stripped of the context that gave it meaning. A voice recording without metadata about the patient’s medication timing was just noise. A facial expression without the conversational context was ambiguous at best, mislead‐ ing at worst. The signal was always there—the data just wasn’t ready to reveal it. That experience, repeated across clinical studies, national-scale health AI deploy‐ ments in Saudi Arabia, and advisory work with the National Academies, has given me a deep conviction: the organizations that will lead in the AI era are not the ones with the most powerful models. They are the ones with the most deliberately architected data. This is precisely the argument that Navnit, Kien, Srikanth, and Harsha make in AI-Ready Data Blueprints, and they make it with a clarity and practical depth that is rare in technical writing. ix

Page 12

What strikes me most about this book is that it’s written based on practical field experience and the examples reflect real customer work that most organizations commonly run into. The authors don’t shy away from the uncomfortable truth that nearly 60% of organizations have not adapted their data strategies for generative AI, even as they pour resources into proofs of concept. They don’t pretend that choosing the right foundation model is the hard part. Instead, they confront the real bottleneck head-on: data preparation matters five to six times more than model selection. Having witnessed this firsthand—from building AI health systems that serve millions to advising governments on national AI strategy—I can tell you that number feels exactly right. The book’s journey mirrors the journey every serious AI practitioner must take. It begins with the foundational question of what makes generative AI fundamentally different from traditional analytics and machine learning, a distinction that too many organizations still underestimate. It then builds systematically through the framework for AI-ready data, the unglamorous but essential work of data wrangling and preparation, the governance and security challenges that become existential at enterprise scale, the architecture of knowledge bases and vector databases, and finally the hard-won wisdom of making AI applications production-ready. I was particularly drawn to the chapters on data governance and production readi‐ ness. In my work serving in a senior AI leadership role for the government of Saudi Arabia, I saw firsthand how governance is not a constraint on innovation but a prerequisite for it. When you are deploying AI systems that touch the healthcare of an entire nation, the questions of data quality, security, compliance, and responsible AI are not afterthoughts. They are the foundation upon which trust is built. The authors understand this deeply, and their treatment of responsible AI principles (fair‐ ness, transparency, accountability, privacy, reliability, and human oversight) reflects a maturity that comes only from real-world implementation experience. What also resonates with me is the book’s vision of where we are headed. The evolution from simple AI assistants to autonomous agents demands a fundamentally different relationship with data. Agents don’t just retrieve information—they reason over it, act on it, and learn from it. The data infrastructure required to support that level of autonomy is qualitatively different from anything we have built before. The authors’ framework for understanding this evolution, along with their practical guidance for preparing for it, is both timely and essential. I have spent my career at the intersection of AI and human well-being, building systems that amplify human ability rather than replace it. The authors share this orientation. Their insistence that data is a strategic asset worthy of deliberate architec‐ ture is not just a technical argument: it is a statement about values. It says that we owe it to the people who will be affected by our AI systems to get the foundations right. x | Foreword

Page 13

Whether you are an executive trying to understand why your generative AI initiative stalled, a data architect redesigning pipelines for the AI era, or a practitioner building your first production RAG system, this book will meet you where you are and take you where you need to go. The blueprint is in your hands. Now it’s time to build. — Ehsan Hoque, PhD Full professor of computer science, University of Rochester Senior AI leadership, government of Saudi Arabia Presidential Early Career Award for Scientists and Engineers (PECASE) MIT Technology Review Innovator Under 35 Foreword | xi

Page 14

(This page has no text content)

Page 15

Preface You’re holding a book that was born from a simple observation: most AI projects don’t fail because of bad models; they fail because of bad data. For the last two years, we have been working alongside organizations of every size across industries as they’ve raced to adopt generative AI. We’ve watched brilliant teams build impressive prototypes, only to see them stall on the road to production. The pattern was always the same. The models worked fine. The prompts were clever. But the data underneath? It wasn’t ready. That pattern was the motivation for writing this book. When ChatGPT launched in November 2022 and reached 100 million users in just 60 days, it set off a wave of excitement and panic across the enterprise world. Suddenly, every boardroom was buzzing with terms like “RAG,” “vector databases,” and “agentic AI.” Companies spun up proofs of concept by the dozen. But as a recent AWS study found, while data leaders overwhelmingly acknowledge the importance of preparing data for generative AI use cases, nearly 60% report that they have not yet made the necessary changes to their company’s data strategies. That gap between knowing data matters and actually making it work is what this book is about. We wrote AI-Ready Data Blueprints for the people in the trenches: the data architects redesigning pipelines for AI workloads, the engineers wrestling with chunking strate‐ gies, the leaders trying to figure out why their AI chatbot keeps hallucinating, and the governance teams wondering how to keep deployments safe and compliant. Whether you’re an executive trying to understand why your GenAI initiative stalled or a hands-on practitioner building your first production retrieval-augmented generation (RAG) system, we wanted to give you something practical—not theoretical, not hype-driven, but grounded in what actually works. xiii

Page 16

What You’ll Find Inside This book follows the journey your data takes—from raw, messy, and scattered to AI-ready, governed, and production-grade. We start by laying out why generative AI demands a fundamentally different approach to data than traditional analytics or machine learning. It’s not just about cleaning up tables anymore. It’s about preserving meaning, modeling relationships, and building systems that can reason, not just retrieve. From there, we walk you through a comprehensive framework for AI-ready data, covering everything from capturing business logic and context to ensuring quality and consistency to managing the security and compliance challenges that come with putting AI into the real world. We explore the nuts and bolts of knowledge bases, vector databases, chunking strategies, and retrieval optimization, because the research is clear: how you prepare your data matters five to six times more than which model you choose. We also confront the challenges you’ll face after you develop a working prototype, delving into topics such as production readiness, automated reasoning, intelligent semantic metadata layers, and the emerging landscape of agentic AI platforms. These aren’t abstract concepts. The insights we provide come from real implementations, including organizations managing quadrillions of files accumulated over decades. Blueprints, architecture diagrams, and sample code are available via the book’s com‐ panion website and GitHub repository. Who This Book Is For If you’ve ever stared at a GenAI demo and thought, “This is amazing—now how do I make it work with our data?” this book is for you. We wrote it for a broad audience: executive leaders, data architects, engineers, AI practitioners, and the domain experts who hold the business knowledge that makes AI actually useful. You don’t need to be a machine learning researcher to get value from these pages. You just need to care about doing AI right. The idea for the book came about after the four of us got together to discuss building data foundations for GenAI for an episode of the podcast Navnit had been hosting on his YouTube channel. We all come from the world of data and AI at AWS. We’ve collectively spent decades helping organizations navigate the messy reality of enterprise data. What unites us is a shared conviction: your data is a strategic asset worthy of deliberate architecture. Get this right, and everything else follows. Get it wrong, and no amount of model sophistication will save you. xiv | Preface

Page 17

We tried to write the book we wished we’d had when this all started—one that’s honest about the challenges, specific about the solutions, and practical enough to use in your production workload. The blueprint is in your hands. Now it’s time to build. Conventions Used in This Book The following typographical conventions are used in this book: Italic Indicates new terms, URLs, email addresses, filenames, and file extensions. Constant width Used for program listings, as well as within paragraphs to refer to program elements such as variable or function names, databases, data types, environment variables, statements, and keywords. This element signifies a general note. This element indicates a warning or caution. Using Code Examples Supplemental material (code examples, exercises, etc.) is available for download at https://oreil.ly/code-samples. If you have a technical question or a problem using the code examples, please send email to support@oreilly.com. This book is here to help you get your job done. In general, if example code is offered with this book, you may use it in your programs and documentation. You do not need to contact us for permission unless you’re reproducing a significant portion of the code. For example, writing a program that uses several chunks of code from this book does not require permission. Selling or distributing examples from O’Reilly books does require permission. Answering a question by citing this book and quoting example code does not require permission. Incorporating a significant Preface | xv

Page 18

amount of example code from this book into your product’s documentation does require permission. We appreciate, but generally do not require, attribution. An attribution usually includes the title, author, publisher, and ISBN. For example: “AI-Ready Data Blue‐ prints by Navnit Shukla, Kien Pham, Srikanth Sopirala, and Harsha Tadiparthi (O’Reilly). Copyright 2026 Navnit Kumar Shukla, AZ25 Lab, Harsha Tadiparthi and Srikanth Sopirala, 979-8-341-63179-3.” If you feel your use of code examples falls outside fair use or the permission given above, feel free to contact us at permissions@oreilly.com. O’Reilly Online Learning For more than 40 years, O’Reilly Media has provided technol‐ ogy and business training, knowledge, and insight to help companies succeed. Our unique network of experts and innovators share their knowledge and expertise through books, articles, and our online learning platform. O’Reilly’s online learning platform gives you on-demand access to live training courses, in-depth learning paths, interactive coding environments, and a vast collection of text and video from O’Reilly and 200+ other publishers. For more information, visit https://oreilly.com. How to Contact Us Please address comments and questions concerning this book to the publisher: O’Reilly Media, Inc. 141 Stony Circle, Suite 195 Santa Rosa, CA 95401 800-889-8969 (in the United States or Canada) 707-827-7019 (international or local) 707-829-0104 (fax) support@oreilly.com https://oreilly.com/about/contact.html We have a web page for this book, where we list errata and any additional informa‐ tion. You can access this page at https://oreil.ly/ai-ready-data-blueprints. For news and information about our books and courses, visit https://oreilly.com. Find us on LinkedIn: https://linkedin.com/company/oreilly-media. Watch us on YouTube: https://youtube.com/oreillymedia. xvi | Preface

Page 19

Acknowledgments The authors wish to express their deepest gratitude to the following people for their support throughout the development of this book: Navnit Shukla This book is, first and foremost, a gift to my family. My first book, Data Wran‐ gling on AWS, was written for my eldest son, Anav. It is a joy to dedicate this work to my second son, Ayansh, who is now 18 months old. To my wife, Anchal, and my sons, Anav and Ayansh: thank you for your unwavering support and for being the light that guided me through the many late nights and early mornings this project required. I would like to extend a special note of gratitude to Sara Hunter for her incredible guidance and support throughout this entire process; her insights were vital to bringing this project to life. I also want to thank the rest of the O’Reilly team— Aaron Black, Elizabeth Faerm, Rachel Wheeler, and Kim Wimpsett—for their editorial expertise. Working alongside my coauthors, Kien, Srikanth, and Harsha, has been a privi‐ lege. I owe them special thanks for the countless hours of debate and our shared commitment to excellence in data architecture. Finally, my thanks go to Long Tran, Thong Do, Robert Fisher, John Giles, and our many reviewers, whose candid feedback kept the technical content grounded and honest. Kien Pham I am deeply grateful to the O’Reilly team—Aaron Black, Sara Hunter, Elizabeth Faerm, Rachel Wheeler, and Kim Wimpsett—whose editorial guidance and patience shaped this book into something far better than what we started with. I owe special thanks to my coauthors—Navnit, Srikanth, and Harsha—for the countless hours of discussion, debate, and shared conviction that enterprise data deserves deliberate architecture. I would also like to thank Long Tran, Thong Do, Robert Fisher, John Giles and many others for their thorough reviews and candid feedback, which kept the technical content honest and grounded in real-world practice. Finally—and most importantly—I thank my family for their unwavering support and understanding throughout the many early mornings and late nights that this project demanded. Srikanth Sopirala I am grateful to the many people who helped bring this book to life. The idea first emerged during a podcast conversation, where a simple discussion sparked a bigger vision for capturing these insights in a more permanent form. Navnit took that early spark and turned it into a reality, providing the encouragement, structure, and accountability needed to transform loose ideas into a finished book. Preface | xvii

Page 20

Thank you to my family and friends for their constant encouragement, to my editor and publishing team for their expertise and guidance, and to the collea‐ gues and readers whose questions, critiques, and conversations shaped these ideas. Your contributions, both seen and unseen, are deeply appreciated. Harsha Tadiparthi To my wife and my little girl— You are the heart of everything I do. This book was written in stolen hours, tucked between bedtime stories and building-block towers, and it would not exist without your unwavering support. To my wife, thank you for holding our world together whenever I disappeared into these pages. To my three-year-old daughter, who has no idea what Daddy has been typing, but whose laughter made every word worth writing. This one is for you both. xviii | Preface